MIME-Version: 1.0
References: <bs7n95ikt65p$.dlg@nililand.de> <4E5B6BF7.80208@morningstar2.co.uk>
Content-Type: text/plain; charset=ISO-8859-1
Message-ID:  <CANQYN6w2ZrLzPY4UqJSwFU_A+Vfw=ApeYgPWjg1UVYTv-k-rew@mail.gmail.com>
Date:         Mon, 29 Aug 2011 10:10:39 -0400
Reply-To: Mailing list for the LaTeX3 project
              <LATEX-L@listserv.uni-heidelberg.de>
Sender: Mailing list for the LaTeX3 project <LATEX-L@listserv.uni-heidelberg.de>
From: Bruno Le Floch <blflatex@GMAIL.COM>
Subject: Re: Removing braces in a token list
To: LATEX-L@listserv.uni-heidelberg.de
In-Reply-To:  <4E5B6BF7.80208@morningstar2.co.uk>
Precedence: list
Status: R

> The is no pre-build 'strip_braces' function at present. You don't say if
> we are talking one set of braces or multiple sets. For the former case,
> a simple 'do nothing' loops would be easiest:
>
>   \tl_set:Nx \l_temp_tl
>     { \tl_map_function:NN \l_temp_tl \exp_not:n }

Agreed with Joseph. If you need to recursively remove braces, and
don't care about speed, the following should work (unless I screwed
up), and it doesn't remove spaces, although it easily could.

\cs_new:Npn \tl_ungroup_all:n
  {
    \tex_romannumeral:D
    \tl_act_aux:NNNnn
      \tl_act_ungroup_normal:nN
      \tl_act_ungroup_group:nn
      \tl_act_ungroup_space:n
      { }
  }
\cs_new:Npn \tl_act_ungroup_space:n #1
  { \tl_act_output:n {~} }
\cs_new:Npn \tl_act_ungroup_normal:nN #1 #2
  { \tl_act_output:n {#2} }
\cs_new:Npn \tl_act_ungroup_group:nn #1
  {
    \tl_act_group_recurse:Nnn
      \tl_act_output:n
      { \tl_ungroup_all:n }
  }

Cheers,
Bruno